Study design and patient selection
The trial design and results of the PIROUETTE trial (Clinicaltrials.gov NCT02932566) have been published previously (12, 14). In brief, between March 7, 2017, to December 19, 2018, 94 patients with HFpEF and myocardial fibrosis were randomised to pirfenidone or placebo for 52-weeks. Eligibility requirements included patients ≥ 40 years of age, symptoms and signs of heart failure, left ventricular ejection fraction of ≥ 45%, and elevated natriuretic peptides (brain natriuretic peptide (BNP) ≥100pg/ml or N-terminal pro-B-type natriuretic peptide (NT-proBNP) ≥300pg/ml; or BNP ≥300pg/ml or NT-proBNP ≥900pg/ml if atrial fibrillation (AF) present). Eligible patients underwent CMR and those with evidence of myocardial fibrosis, defined as an ECV of 27% or higher, were randomised in a 1:1 ratio to treatment with either pirfenidone or matching placebo for 52 weeks using block randomisation, stratified by sex. Key exclusion criteria included alternative causes of patients’ symptoms such as significant pulmonary disease, anaemia, or obesity; pericardial constriction, hypertrophic cardiomyopathy or infiltrative cardiomyopathy; and contraindication to magnetic resonance imaging. The primary outcome was change in myocardial fibrosis, measured using CMR ECV, from baseline to 52-weeks.
The trial was sponsored by Manchester University NHS Foundation Trust. Trial management, independent data management and independent statistical analyses were performed by Liverpool Clinical Trials Centre, a United Kingdom Clinical Research Collaboration fully-registered Clinical Trials Unit. The study protocol was approved by a research ethics committee and trial conduct was overseen by a trial steering committee. Patients were identified at six UK hospitals. Study visits took place at Manchester University NHS Foundation Trust. All patients provided written informed consent.
Study procedures and analysis
The protocol, trial procedures, analysis methods and outcome measurements have been described previously (12, 14). In brief, CMR, echocardiography, electrocardiography, 6-minute walk test, laboratory tests and the Kansas City Cardiomyopathy Questionnaire (KCCQ) were performed at baseline and repeated after 52-weeks of treatment. 31Phosphorous magnetic resonance spectroscopy (31P-MRS), was performed at baseline and 52-weeks in a subset of patients as part of a sub-study (n=60).
Myocardial ECV was calculated from basal and mid left ventricular (LV) short axis T1 maps (MOdified Look-Locker Inversion recovery [MOLLI]), acquired before and 15-minutes following gadolinium contrast (0.15 mmol/kg of gadoterate meglumine), as: ECV = (1 – haematocrit) x [ΔR1myocardium] / [ΔR1bloodpool], where ΔR1 is the difference in relaxation rates (1 / T1) between pre- and post-contrast (4). Haematocrit was measured on the same day as CMR scanning. Absolute myocardial extracellular matrix (ECM) volume was calculated as the product of LV myocardial volume (LV mass divided by the specific gravity of myocardium [1.05 g/ml]) and ECV. Absolute myocardial cellular volume was calculated as the product of LV myocardial volume (LV mass divided by the specific gravity of myocardium [1.05 g/ml]) and (1 – ECV). Further details can be found in the trial protocol paper (14).
Analysis was conducted on an intention to treat basis, including all randomised patients retained in their randomised treatment groups. Continuous data are presented as mean ± standard deviation (SD) or as median (interquartile range (IQR)), as appropriate. Categorical data are presented as counts and percentages. Correlation analyses were used to assess associations between change in ECV (week 52 value minus baseline value), and change in selected secondary outcome variables that reflected cardiac mechanical and electrical function, circulating biomarkers, and functional status. Pearson’s or Spearman’s correlation coefficients were used as appropriate following Shapiro-Wilk testing for normality. Analyses were performed in Stata (Version 14.0, StataCorp, College Station, TX) and SAS (Version 9.4, SAS Institute, Inc; Cary,NC).
Mediation analyses were conducted in order to determine whether changes in myocardial fibrosis, measured using ECV and absolute ECM volume, and changes in myocardial cellular volume (the mediator variables) following antifibrotic therapy, caused changes in cardiovascular structure and function, circulating biomarkers and functional status (the outcome variables).
Mediation analyses were conducted using the Baron and Kenny approach (15), under a structural equation modelling (SEM) framework (Stata package medsem), in order to estimate the average causal mediation effect (ACME) of each mediator, adjusting for confounders (baseline covariates that predicted both the mediator and outcome at the 10% significance level). For each potential outcome, the analysis was only performed if both of the following conditions were satisfied:
The antifibrotic therapy had a significant effect on the outcome at the 5% level (i.e. p<0.05 for the treatment effect in an analysis of covariance (ANCOVA) model of the outcome, adjusting for treatment, sex and baseline value of the outcome variable).
The antifibrotic therapy had a significant effect on the mediator variables (i.e. ECV, absolute ECM volume and myocardial cellular volume) at the 10% level (i.e. p<0.1 for the treatment effect in the ANCOVA models of ECV, absolute ECM volume and myocardial cellular volume, adjusting for treatment, sex and baseline values of the mediator variables).
The outcome variables for this study were all secondary outcome measurements in the main PIROUETTE trial. The outcome variables reflect cardiovascular structure and function, circulating biomarkers and functional status. They were selected for use in this study because they are variables that, based on the published literature and clinical judgment, are associated with adverse outcome in HFpEF or were hypothesised to be impacted by myocardial fibrosis. The selection of the outcome variables was prospective i.e. outcome variables were selected before data lock for the trial had occurred, thus before the results of the trial were known, and were prespecified in an ‘Additional Statistical Analysis Plan’ that was written before data lock.